Changes in the availability of source data for the CE Plant Cost Index periodically necessitate updates to how the index is calculated
The Chemical Engineering Plant Cost Index (CEPCI) is a widely used tool for comparing plant construction costs in the chemical process industries (CPI) between different time periods since 1963. Over its history, the underlying details of the index have been revised and adjusted several times. For a discussion of these changes, readers are directed to Chem. Eng., January 2002, pp. 62–69. This article aims to explain the most recent change that has been made to the CEPCI.
The CE Plant Cost Index is actually a composite that consists of four sub-indices: Equipment; Construction Labor; Buildings; and Engineering and Supervision (see Table 1). The Equipment subindex is itself a composite of seven other indices, each focusing on a particular category of process equipment. The component indexes are compiled, and weighting factors are applied to make up the Equipment subindex. The three other subindexes are compiled independently from appropriate inputs. Information on these four indexes has been reported running back to 1947.
After weighting and normalizing the values, the four sub-indexes are combined to constitute the CEPCI. For each year in the past, an annual value for the CEPCI is calculated based on the arithmetic mean of 12 monthly values for a given year. CEPCI data from previous years can be found and downloaded at www.chemengonline.com/pci.
Table 1. CE Plant Cost Index components
- Heat exchangers & tanks
- Process machinery
- Pipe, valves & fittings
- Process instruments
- Pumps & compressors
- Electrical equipment
- Structural supports & misc.
Engineering & supervision
As many users of the index know, the CEPCI uses specifically selected producer price index (PPI) data and other data from the U.S. Bureau of Labor Statistics (BLS; Washington, D.C.; www.bls.gov), which is a part of the U.S. Department of Labor. Relevant PPI information reported by the BLS is collected by Chemical Engineering editors and used to calculate the monthly changes to the CEPCI. BLS says it is the “principal Federal agency responsible for measuring labor market activity, working conditions, and price changes in the economy.” It collects and analyzes many types of economic information to assist decision-making in the public and private sectors.
A portion of the input data for the CEPCI is derived from the BLS’s PPI data. As the BLS defines them, producer price indexes “track the average change in net transaction prices that domestic producers in the mining, manufacturing, agriculture and forestry sectors, as well as certain service industries, receive for the products they make and sell.” BLS economists and analysts review a statistically chosen representative sample of price quotations in business transactions to build its set of PPIs. Overall, There are around 13,000 PPIs covering more than 600 industries.
A carefully selected group of 41 PPIs is used as raw input to the CEPCI. The set represents a diverse set of materials and components that go into the construction of a CPI facility, including steel plates, concrete pipe, tanks, fans, piping and many others. In addition, the CEPCI uses 12 labor-cost indexes, which are also compiled by BLS.
Making a change
Periodically, the BLS makes changes to the data that they collect and report, some of which necessitate changes in how the CEPCI is calculated. The Chemical Engineering editorial team wanted to make CEreaders aware of one such change that is reflected in this month’s CEPCI (see www.chemengonline.com/pci).
Since the advent of the CEPCI in 1963, one of the inputs has been a PPI data series entitled “Chemical Manufacturing Machinery, Equipment and Parts.” For the purpose of brevity in this article, we will call this data series “CMMEP.”
In autumn of 2015, BLS ceased reporting data for that series, because the information that could be collected did not meet the requirements BLS has established for reporting a data series, including the requirement for a minimum number of reporting units. In some cases, a given PPI may not be reported for several months because the information available does not meet BLS guidelines for reporting, only to return later once available information again meets the reporting criteria. So the absence of data for a particular PPI in a given month is not necessarily a reason to change anything in how the CEPCI is calculated.
However, in the case being discussed in this article, it became clear over the next several months that the CMMEP series would be discontinued permanently, and the series would be subsumed by a broader PPI data series known as “Miscellaneous Special Industry Machinery” (MSIM).
Once the CMMEP was discontinued permanently, the CE editorial team began conversations with BLS economists to inform the magazine’s efforts to determine whether or not the MSIM should be included in the CEPCI as a substitute for the now-discontinued CMMEP. During this time of evaluation, we continued to use the last available data for the now discontinued series to continue to calculate the CEPCI.
CE editorial staff observed the trends in the broader MSIM data series over the past 10 years and compared it to that of the discontinued CMMEP series over the same time period. Unfortunately, the MSIM series did not closely follow the same trends observed in the narrower CMMEP series over the period between 2005 and 2015, indicating that the MSIM would not necessarily be a suitable direct replacement as an input to the CEPCI.
After further consultation* with BLS staff economists and examination of the scope of the MSIM, it was noted that part of the reason for the lack of similarity between the two data series seemed to be that the MSIM included equipment and processes for the semiconductor manufacturing industry. The ways in which specialized equipment has changed for the semiconductor industry sector have been different from how specialized equipment has changed in the CPI over the past 20 years, leading to situation where the data from two series have decoupled and diverged.
The CE editorial team then worked with BLS economists to identify an alternate data series that more closely tracked the data from the discontinued CMMEP series. One series that emerged as a possibly good candidate for replacing the discontinued data series is a PPI Industry series known as “All Other Industrial Machinery” (AOIM).
The AOIM data series differs slightly from the other inputs to the CEPCI, in that it is an industry-based PPI, rather than a commodity-based PPI, as the other CEPCI inputs are. However, economists generally agree that it is possible for the two types of data series to be used together in an aggregate index, such as the CEPCI.
Over the past 12 years, the AOIM data series has tracked much more closely to the discontinued CMMEP series. The similarities can be observed by plotting the values of two data series over time (see Figure 1).
Starting with the March 2018 edition of CEPCI, we began using the monthly changes observed in the AOIM PPI index to adjust the historical data from the discontinued PPI data series. This entails tracking the AOIM and observing the month-to-month changes in that index. These changes, expressed as a percentage, will be applied to the last available value from the discontinued series to derive a new value to input to the CEPCI. This new value will be used to calculate the monthly CEPCI.
In this way, the continuity of the CEPCI can be preserved, while taking into account changes to relevant special industry equipment that are being observed in the PPIs.
Over time, this process will improve the accuracy and timeliness of the CEPCI’s Equipment subindex as well as the overall CEPCI. We will continue to monitor the input data from the BLS in case further changes are necessary.
The CE editors thank the users of the CEPCI for their continued support and attention.
The author would like to thank two individuals who provided their time and expertise in support of this article. Jayson Pollack is a supervisory economist at BLS who graciously spent time on several phone calls to explain how BLS produces the PPI and to identify information that could be used for the CEPCI. John Hollmann is a cost engineering consultant and member of Chemical Engineering’s advisory board who assisted in outlining a strategy to address the discontinued data series. Thank you both very much for your help.
*Editor’s note: Conversations between CE staff and BLS economists explored background information on the PPI data series. BLS did not act in an advisory capacity.
Scott Jenkins is senior editor at Chemical Engineering magazine (40 Wall Street, 50th floor, New York, NY 10005; Phone: 917-370-3073; Email: firstname.lastname@example.org). Before joining the staff of Chemical Engineering in 2009, Jenkins worked in several communications-related positions in the biotechnology and research policy sectors. He has also worked in research and development in the biomedical and chemical fields. Jenkins holds a bachelor’s degree from Colgate University and a master’s degree in organic chemistry from the University of North Carolina at Chapel Hill.
The preliminary value for the February 2018 CE Plant Cost Index (CEPCI; most recent available) increased compared to the previous…
The preliminary value for the January 2018 CE Plant Cost Index (CEPCI; most recent available) increased compared to the previous…
The preliminary value for the December CE Plant Cost Index (CEPCI; most recent available) decreased compared to the previous month’s…
The preliminary value for the November CE Plant Cost Index (CEPCI; most recent available) decreased compared to the previous month’s…
The preliminary value for the October CE Plant Cost Index (CEPCI; most recent available) increased compared to the previous month’s…